infinitecuriosity / reports_i_have_written

These are examples of the quality of graphic design I bring to report writing. One of my professors at Northwestern University said the Wine Data Analysis was so well done that it could be presented to the CEO and Board of Directors. The quality of graphic design means a lot to me, and these examples highlight what can be done in terms of excellent graphic design when applied to data science.

• 👋 Hi, I’m @InfiniteCuriosity! My name is Russ Conte, and I thoroughly enjoy doing data science. It's one of the things I do for fun and play! The whole field is fascinating to me, and I really do have infinite curiosity about methods and solutions that can provide guidance to problems and opportunities.

• 👀 I’m interested in applying data science to actual business issues. My particular skills are around applying data to help people get the best results. This is incredibly rewarding work. Some people call this talent analytics, but at the end of the day it is using the enormous power of predictive analytics to help people live better lives and reach their goals. There are very few things in life more rewarding than reaching a goal, and data science is an amazing tool to help an individual or team reach their goal.

• 🌱 I’m currently creating new solutions advanced R (more on that in a moment), Shiny and Python. The advanced R I'm working on are a set of tools I've developed that are built on top of more common models. Using multiple linear regression as the easiest example, I've developed a template that returns the optimal solution given a tidy data set, and y-hat in the far right column. In addition to the normal solution method, my solution tests all combinations of columns, using all powers from 1/10 to 10, and also includes polynomial values. The solution splits the data into 3 groups (train, test and validation) and returns the model with the highest adjusted r-squared and the model with the lowest MSE, so the results are reproducible, and these results consistently outperform the standard results. These solutions consistently outperform the standard solution methods with all three data sets at the same time (train, test and validation), and this is before feature engineering! Imagine the great results that would come when applying feature engineering to this type of solution method! Apply the same method to many other common solution systems, (such as trees, random forests, Linear Discriminant Analysis, Quadratic Discriminant Analysis, ridge and lasso methods, support vector machines) and this solution method becomes very powerful.

I've also taken time to learn Shiny, which I really enjoy. One of the other areas I'm interested in is graphic design. I'm the long time President of the Chicago Apple User Group, and many of my friends in the group are designers, including a few who are world famous. I use those same amazing design skills to make interactive Shiny apps, and that is a lot of fun for me! Data visualization is a lot of fun, and it's amazing to see the patterns in the data. I'm very comfortable using ggplot2, and I have experience using MatPlotLib and Tableau for data visualizations.

The tidyverse is wonderful to me, so I do a good amount of work using tidyverse. I love how a couple lines of very well written code return a solution, and it's so extremely easy for a reader to follow. My next goal is to get really good at tidymodels.

Recently I've been working to improve my Python skills. I have some experience working with supervised solutions, and my next goal is to master unsupervised solutions using Python (deep learning, Keras, etc.)

• 💞️ I’m looking to collaborate on anything regarding improving talent analytics methods or results, or anything that shows how data science can improve actual business results. I've read many reports where data science teams cannot necessarily provide financial justification for their work. I can show such results, and in fact that's required because of my experience. This means I've got experience helping people and teams demonstrate significantly better results in the real world. My experience has been around recruiting, including identifying and selecting individuals. Several of these methods have set company records for the results they have achieved, but there is much more that can be done! One solution was to reduce employee accidents and that customer went 465 days without an accident using the solutions we developed - that is the best safety result for that company among their 19 locations around the world. Another customer needed 50 temp-to-hire employees. We developed a solution that resulted in all 50 candidates successfully completing the process, without any issues or problems. I'm not aware of any other time 50 employees in a row have been hired without any issues. Those are examples of what data science can do to improve results with talent, and deliver real solutions to businesses.

• 📫 How to reach me. My email address is [email protected], or you can call or text me at [phone].